CHIP 10.5: Agile Iterations (Final Project)
Iteration 3 instructions will be added later.
Dates and point values are subject to change. Changes will be announced on Ed, and Gradescope is the authority for due dates.
CHIP 10.5, Agile Iterations, is the final project for CS 169A: your team of four will develop a SaaS application over three Agile iterations, practicing everything the course has covered — user stories, BDD/TDD, branches and pull requests, continuous integration, code review, and deployment. The project category is worth 20% of your course grade.
The project is documented in three places on this site, plus the instructions:
| Page | What’s on it |
|---|---|
| Instructions | The assignment itself: intro, Overview and Grading (the full rubrics), Iteration 0 setup, and the spec for each iteration. |
| This page | Links, references, and the full timeline. Start here. |
| AI Tools and Use | What AI tools your team may use, what you must disclose, and what is off limits. |
| Scoring and Rubrics | How the 150 points are split across iterations and deliverables, with the Gradescope assignments and due dates. |
CHIP 10.5 Guide (Google Doc)
The project guide is a Google Doc that doubles as your team’s planning document and per-iteration checklist: read the guide or copy it to your Drive. Do the copy once per team, rename it to include your team number, and share it as viewable to anyone at Berkeley so you can link it in your Gradescope submissions. Staff updates appear in the original, not your copy.
Links
| Resource | Link |
|---|---|
| Instructions | 10.5 CHIPS: Agile Iterations (2026) |
| CHIP 10.5 Guide (planning doc) | Preview · Copy to your Drive (once per team) |
| Team repository | github.com/cs169/fa26-team-XX-chip-10.5 (XX is your team number) — coming soon |
| Public starter code | saasbook/hw-agile-iterations |
| GitHub project board template | “[TEMPLATE] 10.5 Backlog Template” in the course GitHub organization |
| Gradescope | Course page — four assignments per iteration, linked on Scoring and Rubrics |
| Teammates | teammatesv4.appspot.com — peer review and self-assessment surveys, released at the end of each iteration |
| bCourses | Course page |
| Ed | Team chat rooms — one private room per team |
| Deployment | https://team-XX.cs169.mball.co — student deployment guide coming soon |
References
Material from the course you will lean on most during the project:
- Textbook: §7 — Requirements: BDD and User Stories, §8 — Software Maintenance: Enhancing Legacy Software Using Refactoring and Agile Methods, and §10 — Project Management: Scrum, Pairing, and Version Control (especially §10.1–10.4 and §10.6).
- CHIPS you’ve done: CHIPS 7.7 (BDD with Cucumber) and CHIPS 8.5 (RSpec on Rails) are the workflow the project expects.
- Tools: GitHub Issues, branches for issues, pull request reviews, GitHub Actions, SimpleCov (code coverage), Cucumber and Capybara.
- Commit messages: prefix each commit with its
conventional commit type,
e.g.
[feat] add representative search.
How it works
Your team works out of a shared repository in the course GitHub organization, with the app deployed throughout — the deployed app is what gets graded. Iteration 0 is setup (everyone can run the app; CI, credentials, deployment, and the issue tracker are in place). It has no deliverable of its own, but Iteration 1 cannot be graded until it is done. Iterations 1, 2, and 3 each have the same four deliverables:
| Deliverable | Who | What |
|---|---|---|
| A. Planning meeting | Team | Early in the iteration: the stories you created as GitHub issues from the spec, with points and owners, plus your meeting notes. In Iteration 3, this is where you propose your team’s two features. |
| B. Code submission | Team | Your repository at the end of the iteration, submitted through Gradescope’s GitHub connection. |
| C. Video + Links Submission | Team | The bulk of the team grade: a short demo video (2–4 minutes in Iteration 1, 2–5 after), the feature work, team work (reviewed PRs, issues, CI), testing (coverage, accessibility, test cases), and how the team used generative AI. |
| D. Retrospective | Individual | Your retro board and action items, the Teammates peer review and self-assessment, your own GitHub contributions, and your AI chats. |
Within each iteration, the recommended rhythm is:
- Planning meeting. Break the spec into 1-point stories, create a GitHub issue for each, label it with the iteration, and assign it.
- Build. One branch per issue; write the Cucumber scenario and RSpec examples first, then the code. Open a pull request when the story is done.
- Review and merge. A different teammate reviews every pull request. CI must be green before merging; merging closes the issue.
- Deploy continuously. Push to your deployment every time
mainchanges so you are never surprised at the deadline. - Standups. A few minutes most days: what did you finish, what’s next, what’s blocking you.
- Retrospective. Meet as a team, then each person submits their own retrospective and fills out the Teammates surveys.
See Scoring and Rubrics for the points and the Gradescope assignments, and Overview and Grading in the instructions for how each deliverable is graded.
Timeline
| Date | Milestone | Late deadline |
|---|---|---|
| Wed Oct 28 | CHIP 10.5 released | Â |
| Mon Nov 2 | Iteration 0/1 begins — project work starts |  |
| Tue Nov 3 | Iteration 1 planning meeting | Â |
| Wed Nov 4 | Iteration 1 planning submission (A) due | Sat Nov 7 |
| Fri Nov 13 | Iteration 1 due — code, video, retrospective | Mon Nov 16 |
| Fri Nov 13 | Iteration 2 begins — planning submission (A) due | Mon Nov 16 |
| Fri Nov 20 | Iteration 2 due — code, video, retrospective | Mon Nov 23 |
| Mon Nov 23 | Iteration 3 begins | Â |
| Tue Nov 24 | Iteration 3 planning submission (A) due | Sun Nov 29 |
| Wed Nov 25 – Fri Nov 27 | Thanksgiving — no classes |  |
| Fri Dec 4 | Iteration 3 due — code, video, retrospective | Sun Dec 6 |
Every Gradescope assignment accepts late submissions until the hard deadline shown on the assignment, two or three days after the due date. There are no further extensions and slip days do not apply — iterations are scheduled team events, like real sprint boundaries.
Teams
Teams of 4 are formed shortly after the add/drop deadline via the team matching survey (required for everyone, even if you’ve already found teammates). You’ll practice working together on the regular CHIPS before the project starts.
Before you start
Finish Module 10 — Scrum, branches, pull requests, and continuous integration are exactly the skills the project exercises. Read the Intro and Overview and Grading pages of the instructions, and the AI Tools and Use page with your team so you agree on how you’ll work before the first planning meeting.
AI Use Policy
This assignment follows the course AI Policy, which will be published on this site before it is needed. Until then, the rule from the Syllabus applies: you must be able to explain everything you submit.
CHIP 10.5 has its own AI rules — agents are allowed, with conditions. Read CHIP 10.5: AI Tools and Use before your first planning meeting.
Running Code and Tests
Run everything below from the root of your clone. If Ruby isn't installed yet, work through Local Development first.
Install the dependencies
bundle install
Run this once after cloning, and again whenever the Gemfile
changes. Prefix the commands below with bundle exec so you run
the same gem versions the autograder does.
Boot the project
Create the database (once, and after any new migration), then start the Rails server:
bin/rails db:migrate
bin/rails serverThe app is then at http://localhost:3000; Control-C stops the server. To try one thing without clicking through the UI, use the console, which loads your app and its models:
bin/rails consoleRun the tests
Run the RSpec suite the way the Gradescope autograder will — get it green locally before you submit:
bundle exec rspec # the whole suite
bundle exec rspec spec/YOUR_FILE_spec.rb # one spec file
bundle exec rspec spec/YOUR_FILE_spec.rb:42 # one example, by line number
bundle exec rspec --format documentation # every example's descriptionEach failure prints the file and line of the failing example along with what it expected — start there rather than at the top of the output.
Run the Cucumber features
Cucumber runs the .feature files — the acceptance tests that
describe the app's behavior from the user's point of view:
bundle exec cucumber # all features
bundle exec cucumber features/YOUR.feature # one feature file
bundle exec cucumber features/YOUR.feature:12 # one scenario, by line